Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Baseline wander removal in ECG

134 views
Skip to first unread message

Corrado Giuliani

unread,
Dec 1, 2010, 4:11:04 AM12/1/10
to
Hi. I'm working on some ECGs for T-wave detection and analysis but i have some problems with the baseline wander removal.
I'm not so good with Matlab,so i'm wondering if anyone knows what should i do to solve this problem.
Thanks.

Corrado Giuliani

unread,
Dec 1, 2010, 4:30:21 AM12/1/10
to
Mmm ok nothing easier. I just applied and high-pass Butterworth filter with 0.5 Hz frequency,filtfiltered all signal and baseline wander magically disappeared.

Corrado Giuliani

unread,
Dec 1, 2010, 4:31:05 AM12/1/10
to

Wayne King

unread,
Dec 1, 2010, 6:08:04 AM12/1/10
to
"Corrado Giuliani" <cioch...@yahoo.it> wrote in message <id54ko$jli$1...@fred.mathworks.com>...

> Mmm ok nothing easier. I just applied and high-pass Butterworth filter with 0.5 Hz frequency,filtfiltered all signal and baseline wander magically disappeared.

Hi Corrado, I'm sure that worked because the baseline drift is DC and you applied a high pass filter. You can also try to use detrend() and see if that does as well as your highpass filter.

y = detrend(x); % removes the best linear fit
y = detrend(x,'constant'); % makes the data zero mean

Wayne

Manas Savkoor

unread,
Mar 1, 2013, 6:07:07 AM3/1/13
to
"Wayne King" wrote in message <id5aak$oci$1...@fred.mathworks.com>...
Hi, This is based on my reading and project-work which uses ECG signals:
1. ECG signal after all processing should have content from 0.01 to 40/100/150 Hz.
2. Simple High pass filtering leads to loss of data! The solution is to find cmfs or dwt and then find the residual! That is your baseline wander!

Regards,
Manas
0 new messages